The use of simulation and optimization techniques within the supply chain environment is now commonplace. There is a myriad of software tools (many of which are baked into traditional ERP/MRP solutions or offered as add-ons). The concepts that simulation tools provide can offer many benefits such as improved accuracy, speed and ease of use. These can deliver robust models that can better equip strategists and business decision makers.
However, simulation and optimization is not perfect and like any tool or process have some things which they are not good at.
1. Your initial data is accurate right?
The key to simulation tools is that they require data – if that data lacks accuracy then the results obtained from any modelling will be questioned. For example if your looking to run a simulation to better understand long term costs and your utilizing data that has inaccurate cost estimates then your doomed to failure. It may sound obvious but simulation is not a cure for poor ERP data.
2. How robust is your demand profile?
Optimization is usually centered on a estimated demand over a period of time. You can modify the estimated demand profile and run various scenarios comparing the effect on the recommended solution; however, optimization can struggle where demand or other inputs are highly variable.
Where inputs, are not constant, but more dynamic some tools may require more manual intervention and review. This can be more important where the estimated demand required is more granular i.e. based on a daily (or greater) interval.
3. How did you come up with that?
Optimization also has a tendency to be a closed or “black box” method. This means accepting inputs, figuring numbers and offering a solution but where the calculation itself remains a mystery – the system just does it.
Due to the complexity of such systems, it is not always easy for the operator to fully understand the relationships of the different factors, how all of the inputs, variables and data work as unit. This can often be further hampered by the resellers themselves who fail to fully explain how the system defines results.
The significance of this problem can vary depending on the accuracy of results coupled with the intended audience. Imagine giving your presentation to the CEO and he asks “how did you come up with that” and you can’t provide an accurate answer!
4. It’s a system not a crystal ball!
Whilst the benefits of simulation include the ability to better understand the impact of dynamic events, systems will never predict everything. There are likely to be various issues that every system will struggle to predict – for example consider how obsolescence can be reviewed in long term planning or how events such as the Icelandic volcano eruptions in 2009 which caused huge issues for international flights could’ve been predicted. Remember it’s a tool not a soothsayer!
Summary
The key to any system is to know its limitations – the many benefits of simulation overcome its few shortcomings but businesses must be aware that it is heavily reliant on the data and variables that it is fed with. Simulation and optimization tools will not fix any existing issues that the business has with these and it will not completely plug the gap in operational knowledge – results will still need to be reviewed, interpreted and iterated to get maximum benefit.